Research project In order to provide individuals instruments for managing stress and cognition in stress-related disorders, an AI-based digital companion, or coach is being developed in a collaboration between Umeå University, Karlstad University and the stress rehabilitation clinic at the Västerbotten County Council.
The purpose with the project is to develop AI-based person-tailored decision support in the form of a digital coach for managing stress, fatigue, and tiredness, and to provide cognitive support in the process of going back to work when having a stress-related disease.
Stress-related diseases are increasing in Sweden and is connected to long periods of abscence from work, with high cost for both the individual and the society. To find methods for finding balance in life and facilitate the process back to work is highly important.
The purpose with the project is to develop AI-based person-tailored decision support in the form of a digital coach for managing stress, fatigue, and tiredness, and to provide cognitive support in the process of going back to work when having a stress-related disease. The decision support will use AI-methods for collaborating with the person from the perspective of the individual's goals, needs, motivation and interests in daily life and work. The decision support will be evidence-based and pro- actively coach the person in daily activities that aims at rehabilitation, e.g. physical and cognitive exercise, and balancing rest and activity.
The digital coach will use partly new developed AI-methods based on machine learning and argumentation logics. To give a holistic support to the person, also earlier developed software and AI-methods will be integrated, e.g. a cognitive exercise program that has shown good results on cognitive functions in different patient groups. Patients in the target group will be involved in a participatory design process to shape the behaviour of the coach, and focus groups will be creating personal smart objects in SoftLab, Sliperiet (Umeå University) as part of the interaction design.
The results will be useful for also other groups of patients and citizens.